Feedback analysis of climate change simulations

Kurzfassung

The climate sensitivity parameter that describes the change in surface temperature due to a unit change in radiative forcing has long been assumed to be constant. However, recent studies found that the climate sensitivity parameter varies, not only amongst models for the same forcing but also within the same model where it may strongly depend on the strength and on the type of the applied radiative forcing. By means of the “Partial Radiative Perturbation”-method (PRP-method), a complete feedback analysis of CO2 driven climate change simulations is performed to identify the individual feedback processes which are responsible for the variation in climate sensitivity parameter. To include all components of the feedback analysis, the stratospheric temperature feedback is introduced in this work. It describes the stratospheric temperature change due to a radiative forcing. This feedback is found to be weakly positive. The combination of the stratospheric temperature feedback and the instantaneous radiative forcing allows to approximate the stratosphere adjusted radiative forcing which is known to be a better climate predictor than the instantaneous forcing. In a set of CO2 driven equilibrium climate change simulations, the water vapour, the cloud and the stratospheric temperature feedback are found to vary the most under increasing radiative forcing. Hence, the interplay between these three feedback processes causes an increase of the climate sensitivity parameter when the atmospheric carbon dioxide concentration is quadrupled in comparison to a doubling of the CO2 concentration. For climate change simulations with a small CO2 radiative forcing, it was not possible to identify the feedback processes which are responsible for a varying climate sensitivity parameter. Thus, forcings must be sufficiently large to establish significant differences of feedbacks that are interpretable to explain differences in climate sensitivities. Feedbacks of CO2 driven simulations with and without interactively coupled atmospheric chemistry are also compared. Only the stratospheric temperature feedback differs significantly among these simulation experiments. For the simulation without interactively coupled chemistry, the stratospheric temperature feedback is considerably larger than for the simulation with interactively coupled chemistry where the trace gases could adjust to the radiative perturbation. The change in ozone is found to be responsible for the difference between these simulations. Ozone changes to a CO2 radiative forcing causes a negative feedback, which reduces the stratospheric temperature feedback, when the reaction of the atmospheric chemistry to the CO2 perturbation is included. Moreover, the strengths and weaknesses of the PRP-method are investigated. This method is only suitable for calculating independent feedbacks and to yield a balance of radiative forcing and feedbacks, if the forward and backward PRP calculations are combined. If only the forward or the backward PRP calculation is considered, interactions between feedbacks occur, which render the separation of the climate response into individual feedbacks as unpracticable. In particular, the water vapour and the lapse rate feedback as well as the water vapour and the cloud feedback show large overlapping effects. These overlapping effects are completely erased when forward and backward PRP calculations are combined.